Towards a Comprehensive BPMN Extension for Modeling IoT-Aware Processes in Business Process Models [original]

Towards a Comprehensive BPMN Extension for

Modeling IoT-Aware Processes in Business

Process Models

Yusuf Kirikkayis, Florian Gallik, and Manfred Reichert

Institute of Databases and Information Systems, Ulm University, Germany

{yusuf.kirikkayis,florian-1.gallik, manfred.reichert}@uni-ulm.de

Abstract. Internet of Thing (IoT) devices enable the collection and ex-

change of data over the Internet, whereas Business Process Management

(BPM) is concerned with the analysis, discovery, implementation, execu-

tion, monitoring, and evolution of business processes. By enriching BPM

systems with IoT capabilities, data from the real world can be captured

and utilized during process execution in order to improve online process

monitoring and data-driven decision making. Furthermore, this integra-

tion fosters prescriptive process monitoring, e.g., by enabling IoT-driven

process adaptions when deviations between the digital process and the

one actually happening in the real world occur. As a prerequisite for

exploiting these benefits, IoT-related aspects of business processes need

to be modeled. To enable the use of sensors, actuators, and other IoT

objects in combination with process models, we introduce a BPMN 2.0

extension with IoT-related artifacts and events. We provide a first eval-

uation of this extension by applying it in two case studies for modeling

of IoT-aware processes.

Keywords: BPMN ·BPM·Internet of Things ·IoT in BPM ·Sensors

·Actuators

1 Introduction

As electronic components have become smaller, more powerful, and less expen-

sive, the Internet of Things (IoT) has received an upswing in recent years [1].

Many embedded components are equipped with sensors and actuators that en-

able collection of environmental data (sensors) as well as physical responses to

specific events (actuators) [2]. IoT components can be embedded in everyday ob-

jects such as washing machines, refrigerators, vehicles, cell phones, or wearable

devices. Moreover, they can be found in cyber-physical systems, smart cities,

or smart logistics [3]. IoT refers to a network of physical objects or ”things”

being equipped with sensors, actuators and software to connect them with other

devices and systems over the Internet. Such interconnected devices, in turn,

constitute the basis for exchanging data [2].

While IoT allows capturing and exchanging data about the physical envi-

ronment, BPM enables the analysis, discovery, implementation, execution, mon-

itoring, and evolution of business processes [17]. BPM-enabled processes can

2 Y. Kirikkayis et al.

be further enhanced by sensors, e.g., to measure the fill level of a tank and

eliminate the need for manually performing this task [5]. In general, IoT tech-

nology contributes to make abstract process models real-world-aware and, thus,

to align digital processes with the physical world[15]. Moreover, IoT devices can

be used to automate different types of tasks, which may be physical (moving

a conveyor belt) or digital (sending data or notifying a system) [14]. Further-

more, IoT enhances the monitoring, discovery, and optimization of processes,

which are referred to as IoT-aware processes in the following. An important task

for modeling IoT-aware processes is to properly capture IoT-related aspects.

Modeling a process fosters the understanding of how the process works and al-

lows discovering potential problems (e.g., deadlocks) before process automation.

Finally, already modeled processes can be analyzed, improved, automated and

optimized [17].

There are several languages for modeling business processes, such as Petri

Nets, Event-driven Process Chains (EPC), Role Activity Diagrams, Resource-

Event-Agent (REA), and Business Process Modeling Language (BPML) [16]. A

standardized process modeling language is Business Process Model and Notation

(BPMN) 2.0, which has undergone three releases since 2004 [6]. Modeling IoT-

aware processes with BPMN 2.0 is a complex endeavor, and the resulting model

is difficult to understand due to the potentially ambiguous use of modeling ele-

ments. As a drawback BPMN 2.0 does not allow for the explicit representation

of IoT devices and IoT-related aspects, which aggravates the maintenance and

servicing of IoT-aware processes significantly. In particular, the resulting pro-

cess models lack structure, expressiveness, and flexibility. To overcome the lack

of language elements for modeling IoT aspects, BPMN 2.0 needs to be extended.

Existing BPMN 2.0 extensions for IoT-aware processes are either incomplete or

do not comprehensively cover the required treatment.

In this work, we present a BPMN Extension for IoT-aware processes which

enables the explicit integration of business process models with IoT devices. The

approach supports the modeling of IoT-aware processes in terms of different

views and levels of abstraction.

The remainder of this paper is organized as follows: In Section 2, we sum-

marize the main issues that emerge when modeling IoT-aware processes with

the existing BPMN 2.0 standard. Section 3 discusses existing works. In Section

4, we present our BPMN 2.0 extension for modeling IoT-aware processes and

illustrate it along two case studies. Finally, Section 5 summarizes and discusses

our approach.

2 Problem Statement

In order to properly model the behavior of IoT-aware business processes a multi-

tude of input and output devices may have to be integrated with a process model.

In principle, BPMN 2.0 offers various mechanisms for representing IoT devices.

On one hand, script, service and business rule tasks can be used to represent IoT-

related activities. On the other, resources, data objects or events may be used

Towards a Comprehensive BPMN Extension for IoT 3

to model IoT-involvement [7]. However, when following such a straightforward

approach, no distinction between regular BPMN tasks and IoT-related ones can

be made. The familiarization with such a process may therefore take longer, as

the IoT-related model elements cannot be visually distinguished from standard

BPMN elements. Consequently, the modeled IoT tasks constitute a black box,

i.e., it does not become transparent whether the task refers to a sensor, an actu-

ator, or a service call [7]. On one hand, this aggravates model comprehensibility,

on the other it results in a limited usability and poor maintainability of the

model.

The process model, depicted in Figure 1, deals with the treatment of the

Chronic Obstructive Pulmonary Disease (COPD). COPD describes the obstruc-

tion of the lungs, which hinders the patient’s breathing [7]. First, the patient’s

heart rhythm is checked (1). If necessary, an emergency alarm is triggered (2).

Then the severity of the COPD is assessed (3) and, depending on the outcome

of this assessment, either no treatment, treatment with an oxygen mask (4), or

treatment with an inhaler (5) is administered. Finally, the results of the treat-

ment are analyzed (6) and the patient record is updated accordingly (7).

Check heart

rhythm

Check COPD

severeness

Sound

emergency

alarm

Administer

inhaler

Mild

attack

Administer

oxygen mask

Severe

attack

OK Analyze result of

treatment

Update patient

record

Fig. 1. Example of a process model with IoT aspects. (Adapted from [7])

When using standard BPMN 2.0 elements for modeling the physical (i.e.

IoT-related) tasks of the COPD process (Figure 1), it is unclear, which tasks

are IoT-related and which are not. In addition, it is unclear which sensors and

actuators, respectively, are involved in the processing of the IoT-related tasks.

Instead, it becomes necessary to carefully read and understand the underlying

process model in order to make assumptions whether, for example, a business

rule task refers to a specific sensor or a service task represents an action of an

actuator. Note that this might cause ambiguities due to labeling issues.

When representing sensors in terms of business rule tasks, the involvement

of IoT devices (cf. Figure 1, Activity (1) heart rate sensor) does not become

apparent as well. Finally, there is no visual difference between an IoT-related

Business Rule Task (1 & 3) and a BPMN Business Rule Task (6), or between

an IoT-related Service Task (2) and a BPMN Service Task (7). This aggravates

the comprehension as well as maintenance of the process model.

4 Y. Kirikkayis et al.

3 Existing Approaches

There exist several works that introduce notations, approaches, or language ex-

tensions for representing IoT devices in the context of BPMN 2.0. This section

briefly describes these approaches and discusses them (Table 3).

[1] and [7] model IoT-aware processes with standard BPMN 2.0 elements.

While [1] uses script tasks for integrating both sensors and actuators, [7] uses

business rule tasks for representing sensors and service tasks for actuators. Cheng

et al. [12] extend the BPMN 2.0 standard with a sensor task covering the follow-

ing aspects: sensor device, sensor service, and sensor handler. Another approach

for representing physical entities (e.g., a bottle of milk) in terms of a collapsed

pool is presented in [10]. In particular, for sensing and actuation activities two

new task types are introduced. Sungur et al. [13] explore the properties of wire-

less sensor networks (WSNs). For this purpose, they introduce a WSN Task and

a WSN Pool. The WSN Task has an actionType element consisting of a question

mark, an exclamation mark, and a square. It is used to specify a WSN opera-

tion as a sensing (?), actuating (!), or intermediating operation (˝). A real-world

temperature control scenario is suggested by [4], which enhances existing BPMN

2.0 events with a conditional event, message event, and error event. [8], extends

BPMN 2.0 with a resource called ResourceExtension, included for both human

and non-human resources. The ResourceExtension has some privileges (Resour-

cePrivileges) and types such as RFID, Sensor and Actuator (ResourceTypes).

uBPMN [6] suggests additional elements for Sensor, Reader, Collector, Camera,

and Microphone. Each of these elements is represented by specific task and event

types. In addition, a Smart Object is introduced to represent transmitted data.

In [9], an Industry 4.0 process modeling language (I4PML) extending BPMN

2.0 with the following elements is presented: Cloud app, IoT device, device data,

actuation task, sensing task, human computer interface, and mobility aspect.

Though existing approaches already enable the modeling of various IoT-

driven process scenarios, there remain some gaps or scenarios that cannot be

fully represented. Except for [1] and [7], all other approaches extend BPMN

2.0 with specific IoT elements. While uBPMN only introduces a Start Event,

none of the approaches explores the execution and/or control of an actuator in

combination with an End Event. In addition, none of these approaches allows

for the combined use use of sensors and actuators in the context of a task.

Furthermore, none of the approaches supports responses to an IoT event during

task execution. Another important scenario that cannot be modeled with existing

approaches is the verification of an IoT-driven condition when processing a task.

There is also no concept for representing of IoT-driven processes with different

levels of abstraction in the already existing approaches. Note that modeling IoT-

driven processes with different abstraction levels could enable different views for

various stakeholders (e.g., domain expert, BPMN expert, or IoT expert).

Table 3 provides a systematic summary of the different approaches (with ✓in-

dicating support of the respective feature and ✗expressing missing support). As

can be easily seen, non of the approaches comprehensively covers the treatment

needed for IoT-aware processes.

Towards a Comprehensive BPMN Extension for IoT 5

Sensor

Actuator

Start Event

End Event

Intermediate Event

Condition Element

Physical Entity

IoT Data Object

Abstraction level

Score

Meyer et al.[10] ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ ✗ ✗ 3/10

Sungur et al.[13]. ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ ✗ ✗ 3/10

uBPMN[6]. ✓ ✗ ✗ ✓ ✗ ✗ ✓ ✓ ✗ ✓ ✗ 5/10

Cheng et al[12] ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ 1/10

BPMN4WSN[11] ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ ✗ ✗ 3/10

Suri et al.[8] ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ ✗ 2/10

I4PML[9] ✓ ✓ ✗ ✗ ✗ ✗ ✗ ✗ ✓ ✓ ✗ 4/10

Combining Sensor

and Actuator

React to IoT

within a Task

Table 1. Currently supported IoT elements through the extensions

4 Solution Proposal

To tackle the gaps and problems discussed in Sections 2 and 3 respectively, we

extend BPMN with artifacts and events as shown in Figure 2. Note that these

artifacts and events are not limited to specific use cases or processes, but may be

used in any domain. Due to lack of space, the various elements cannot described

in detail. Instead, we demonstrate the use of selected IoT artifacts and events

along two IoT-aware processes from different domains. Note that all elements

are decorated with a WLAN icon and labeled as “IoT”. In addition, the latter

in the upper left corner indicates the artifact type.

Sensor

Artifact

Actuator

Artifact

Actuator Group

Artifact

Sensor

Group Artifact

Catch

Artifact

Catch Group

Artifact

IoT Start

Event

IoT End

Event

IoT Intermediate

Catch Event

IoT Intermediate

Throw Event:

Object

Artifact

IoT Boundary

Event

S A S S

Fig. 2. Extended BPMN 2.0 elements

6 Y. Kirikkayis et al.

4.1 Case Study 1: Healthcare Process

In Case Study 1, we consider an IoT-enabled measurement process in a medical

facility. The process can be modeled with our extension as shown in Figure 3.

The process begins when the physician registers the treatment with the help of

an RFID scanner. The patient is then transported to the treatment room. In

the treatment room, sensors are used to check whether the patient has arrived.

If this does not happen within 10 minutes, a notification appears on a specific

monitor and the process ends. If the patient has arrived within the specified

time period, a measurement process is started. Afterwards, the measured data is

received, the evaluation is started, and a severity score is calculated. This score

serves as a basis for the next steps. If the score is above 8, an alarm is triggered

with an IoT actuator artifact. If the score is 6 or 7, an indicator light is switched

on. If the score is below 6 nothing happens. Finally, the process ends with a

logout at the RFID sensor.

RFID scanner

activated

Transport

patient to

treatment room

Check if patient

arrived

Start

measurement

process

Show

notification

Calculate

severity score

Turn on indicator

light

Start alarm

6.7

≥ 8

< 6

Logout on RFID

scanner

> 10 min

Notification

screen

Light barrier &

GPS location

Measurement

instruments

Measurement

sensors

Alarm

Indicator

light

Fig. 3. Healthcare process of Case Study 1

4.2 Case Study 2: Production Process

Case Study 2 considers a process that sort workpieces based on their color in a

production setting. This process could be modeled with our extension as shown

in Figure 4. Process execution starts as soon as the light barrier on the conveyor

belt is triggered by an incoming workpiece. The conveyor belt is then started

and remains in operation until the end light barrier is triggered or the weight

on the conveyor belt exceeds 1000 kilograms. A light is then switched on inside

the machine and a color sensor is used to determine the color of the workpiece.

Finally, the workpiece is sorted according to its color.

Towards a Comprehensive BPMN Extension for IoT 7

Light barrier

triggered

Start conveyor

belt

Get Color of

workpiece

Weight on belt

> 1000kg

Stop conveyor

belt

Conveyor belt

End of conveyor

belt reached

Stop conveyor

belt

light color sensor

Sort workpiece by

color

Fig. 4. Sorting process of Case Study 2

As shown, the proposed BPMN 2.0 extensions enable a proper modeling of

the two IoT-aware processes analyzed in the case studies. In particular, both

processes can be modeled more intuitively and specifically compared to the ap-

proaches discusses in Section 2 and 3.

5 Conclusions

This paper presented a BPMN 2.0 extension for modeling IoT-aware processes.

Based on the problem description and a literature review, we identified gaps and

problems of existing BPMN extensions for modeling IoT-aware processes. We

then extended BPMN 2.0 with additional IoT artifacts and IoT-related events

that address the identified gaps. In particular, the added elements enable the

acquisition of physical data with the sensor artifact and the control of actuators

with the actuator artifact. Furthermore, subjects and/or objects can be repre-

sented by IoT objects. IoT conditions, in turn, can be validated during task

processing by the IoT intermediate catch artifacts as well as along the sequence

flow by the IoT intermediate events. All artifacts can be aggregated into corre-

sponding group artifacts to increase the abstraction level. Moreover, the process

start may be triggered by an IoT condition associated with an IoT start event.

In addition, a process end may execute and/or control an actuator with the IoT

end event. Finally, we introduced an IoT boundary event, which allows redirect-

ing the sequence flow based on an IoT condition. We have applied our extension

in two case studies to show how the introduced artifacts and events can be used.

In future work we will perform various experiments and studies with different

users such as BPMN modelers or domain experts to investigate the completeness

of our extension and to study model comprehensibility. Furthermore, we inte-

grate our extension with a process engine, i.e., the framework should support

both the modeling and execution of IoT-aware processes.

8 Y. Kirikkayis et al.

Acknowledgments This work has been funded by the Deutsche Forschungsge-

meinschaft (DFG, German Research Foundation) under project number 449721677.

References

1. Domingos, D., Francisco, M.: Using BPMN to model Internet of Things behavior

within business process. In: International Journal of Project Management, pp. 39–

51. (2017).

2. Janiesch et al.: The Internet of Things Meets Business Process Management: A

Manifesto. In: IEEE Systems, Man, and Cybernetics Magazine 6(4), pp. 34–44.

(2020).

3. Chang, C., Srirama, S., Buyya, R.: Mobile Cloud Business Process Management

System for the Internet of Things: A Survey. In: ACM Computing Surveys 49(4),

pp. 1–42. (2016).

4. Chiu, H., Wang, M.: Extending Event Elements of Business Process Model for In-

ternet of Things. In: IEEE International Conference on Computer and Information

Technology, (2015).

5. Cherrier, S., Deshpande, V.: From BPM to IoT. In: 15th International Conference

on Business Process Management, pp. 310–318. (2017).

6. Alaaeddine, Y., Bauer, C., Saidi, R., Anind, D.: uBPMN: A BPMN extension for

modeling ubiquitous business processes. In: Information and Software Technology,

(2016).

7. Hasi´c, F., Asensio, E.: Executing IoT Processes in BPMN 2.0: Current Support and

Remaining Challenges. In: 13th International Conference on Research Challenges in

Information Science (RCIS), (2019).

8. Suri, K., Gaaloul, W., Cuccuru, A., Gerard, S.: Semantic Framework for Internet

of Things-Aware Business Process Development. In: 26th International Conference

on Enabling Technologies: Infrastructure for Collaborative Enterprises, (2017).

9. Petrasch, R., Hentschke, R.: Process Modeling for Industry 4.0 Applications Towards

an Industry 4.0 Process Modeling Language and Method. In: 13th International

Joint Conference on Computer Science and Software Engineering (JCSSE), (2016).

10. Meyer, S., Ruppe, A., Hilty, L.: The Things of the Internet of Things in BPMN.

In: Conference in Advanced Information Systems Engineering Workshops, (2015).

11. Tranquillini et al.: Process-Based Design and Integration of Wireless Sensor Net-

work Applications. In Business Process Management, (2012).

12. Cheng et al.: Modeling and Deploying IoT-Aware Business Process Applications

in Sensor Networks ,(2019).

13. Sungur et al.: Extending BPMN for Wireless Sensor Networks. In: IEEE 15th

Conference on Business Informatics, (2013).

14. Janiesch, C. et al.: The Internet-of-Things Meets Business Process Management:

Mutual Benefits and Challenges, (2017).

15. Gruhn et al.: BRIBOT: Towards a Service-Based Methodology for Bridging Busi-

ness Processes and IoT Big Data, (2021).

16. Mili, G., Tremblay, G., Jaoude, G., Lefebvre E., Elabed, L., Boussaidi, G.: Business

process modeling languages: Sorting through the alphabet soup. Association for

Computing Machinery (ACM), New York (2010).

17. Dumas, M., La Rosa, M., Mendling, J., Reijers, H.: Fundamentals of Business

Process Management 2nd edn. Springer-Verlag GmbH, Germany (2018).